West End 61 | West End 61 | Getty Images
Companies with hundreds or thousands of employees across functions face a daunting task in the upskilling race. Training employees at scale on generative artificial intelligence is even more of a burden given the technology's novelty, but organizations are doing it nonetheless — not just for the sake of their companies' efficiency and innovation, but also for the long-term success of their employees (after all, according to Microsoft's 2024 Work Trends Index, 66% of leaders will not hire anyone without AI skills).
These organizations that are pushing forward with widespread generational AI training are learning lessons along the way.
Take Synechron, a global IT services and consulting organization. The majority of Synechron's roughly 13,500 team members are AI-enabled, a result of very deliberate training. Because many of its clients operate in regulated environments, Synechron has created a suite of nine secure in-house solutions, one of which is an application similar to ChatGPT called Nexus Chat. Of Synechron employees who don't work at restricted client sites, 84% use Nexus Chat.
“Once we give people access to the tools, we can start to teach them,” says Synechron CTO David Sewell. They started with online courses, teaching beginner-level prompt engineering (how to interact with and get the most out of AI). They also created videos showcasing potential use cases for a range of non-technical roles, like HR and legal, and included a survey to speed up familiarity.
As a sort of pilot, Synechron gave a small group of technical and general employees early access to some of its tools, including Unifai, an AI-powered HR bot that's trained on sensitive HR policies and company data, and has now been adopted by 74% of the workforce.
On the technology side, Sewell says productivity in the software development lifecycle has increased by 39%. Non-technical roles are harder to measure, but Synechron's chief marketing officer Antonia Maneta says, “After just a few months, we can't see how we could have run our business without AI. It's incredible. The changes in productivity have freed up time to work on what's important, what our focus areas are.”
Amara D'Aguilar, chief information officer at financial services company USAA, is planning AI training for her company's 37,000 employees. Her strategy boils down to three main steps:
The first one is centered around governance and risk management. Second, senior leaders are trained on the solution that has passed the governance analysis (the training for senior leaders is primarily done through face-to-face sessions with industry experts). Finally, different types of teams go through educational courses, which vary depending on whether their role is to create technology, protect the organization from risk, or simply use the tools provided.
Hack, train, repeat
To get hands-on experience with generative AI solutions, USAA and other companies are seeing hackathons — events where employees of all kinds can present new ways to use the technology — as an effective tool. A recent USAA employee hackathon drew a record number of participants from both technical and nontechnical teams, with enthusiasm ranging from software to compliance. D'Aguilara says they were able to bring about 55 use case ideas into a controlled environment for testing. “We were surprised at the level of interest in this within the organization,” she adds.
“We got feedback from people who attended the hackathon that they didn't feel fully prepared,” Synechron's Sewell said of the hackathon, adding, “We took the feedback from that event and started creating more training material to educate them. We gave them a bit more space and time to get comfortable with the technology before they were asked to demonstrate efficiencies or benefits.”
Not all workforces are created equal
Terry O'Daniel is head of security at Amplitude, a digital analytics platform whose clients include Atlassian, Under Armour, Walmart, and BeReal. O'Daniel, who previously ran security at Instacart, Netflix, and Salesforce, prioritizes clear guidelines and refinements over aggressive training. “We went through and tried to understand every use case,” he says. “We quickly learned that that wasn't what the business was looking for. They were looking for something very practical, tangible way to solve the problem at hand.”
Still, O'Daniel believes AI evangelism is beneficial, and that much of it comes from his IT and security teams. “We help people understand that we're not the department that says 'no.' We're the department that says, 'try this instead.'” His guidelines, which focus on data privacy and security, intellectual property, and output validation, mean employees can ask questions about the platforms they plan to use, and he often seeks approval from his team before introducing new AI solutions into their workflows.
For example, Amplitude has a corporate subscription to the OpenAI API feed, and O'Daniels ensures employees use it rather than public solutions that share data.
This is possible for Amplitude, which has more than 700 employees, but companies with tens of thousands of employees, as is the case for USAA and Synechron, opt for a more structured solution. Synechron's evangelism method is also more structured, with Ryan Cox, the company's head of AI, visiting many of its offices around the world to identify dedicated employees who can take the message of AI and its training solutions to local audiences.
This shift confirms that, without neglecting responsible use of AI, companies must find what works best for them.
“If we don't think about it, we're going to fall behind,” USAA's D'Aguilar said. “If we don't think about it in a responsible way, we're going to fall even further behind.”
